No. 361 of 500 No. 83 in AI Ethics & Society

More than a Glitch

Confronting Race, Gender, and Ability Bias in Tech

By Meredith Broussard

Published
April 2, 2024
Print length
248 pages
Publisher
The MIT Press
Reading level
Beginner
ISBN-13
978-0262548328
Written for
Policymakers, technologists, students and general readers concerned with fairness in automated decisions.

Synopsis

Meredith Broussard, a data journalist and former software developer, takes aim at the comfortable idea that discriminatory outcomes in technology are glitches that a quick fix will remove. Her argument is that algorithms trained on data produced by an unequal society reproduce that inequality, and that the belief in computational solutions as inherently superior, which she calls technochauvinism, keeps the problem hidden. She builds the case through examples: facial recognition that fails more often on darker skin and has contributed to wrongful arrests, mortgage models that disadvantage Black applicants, predictive policing, automated grading, and medical tools that perform worse for underrepresented groups.

Broussard extends the analysis to gender and disability, examining how databases built on binary categories handle people whose identities do not fit them and where assistive technologies help or fall short. In a chapter drawn from her own breast cancer diagnosis, she examines what an AI reading of her scans could and could not do. Rather than asking only for more inclusive products, she argues for public interest technology, regulatory sandboxes and routine algorithmic auditing, treating software like bridges that require ongoing inspection. Readers gain a clear vocabulary for evaluating automated decisions, concrete cases to cite, and a practical sense of what accountability for algorithms would require.

Scorecard

Editorial score

How it scored

Editorial score: 69.5 / 1005 weighted criteria

  • Influence30%

    21.0/30
  • Authority25%

    19.5/25
  • Practical value20%

    11.0/20
  • Currency15%

    10.5/15
  • Readability10%

    7.5/10
  • Editorial score

    69.5/100

How the rank was computed

Rank No. 361 of 500 comes from a composite of 51.3: observable signals 19.9 of 55 points plus editorial review 31.4 of 45 points.

All 11 inputs
InputPoints
Observable signals19.9/55
Reach (Amazon ratings volume)6.7/16.5
Consensus across best-of lists5.5/13.8
Awards and bestseller evidence0.0/8.3
Reader rating, sample-adjusted5.2/8.3
Amazon sales rank0.8/5.5
Recency1.7/2.8
Editorial review31.4/45
Author authority8.8/11.3
Influence9.5/13.5
Practical value5.0/9.0
Readability3.4/4.5
Currency4.7/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

An accessible, well-argued case that algorithmic bias is structural, not accidental, and that automated systems need ongoing audits.

Strengths

  • Bridges computer science and social science
  • Vivid, well-chosen real-world case studies
  • Concrete remedies such as algorithmic auditing

Keep in mind

Readers skeptical of structural accounts of discrimination may dispute its framing, and its policy proposals are sketched rather than fully developed.

Key takeaways

  1. Bias in automated systems is usually structural: models trained on data from an unequal past reproduce that inequality unless designers actively prevent it.
  2. Technochauvinism, the assumption that computational solutions beat all others, leads institutions to deploy algorithms where human judgment or simpler methods would serve better.
  3. Facial recognition performs worse on darker skin, women, and trans and nonbinary people, making its use in policing especially risky.
  4. Algorithms need continuing audits and maintenance, like physical infrastructure, and some high-stakes uses should be abandoned rather than merely made more inclusive.

Notable facts

About the author

Meredith Broussard

Meredith Broussard is an associate professor at New York University's Arthur L. Carter Journalism Institute and research director of the NYU Alliance for Public Interest Technology. A former features editor at The Philadelphia Inquirer, she worked as a software developer at AT&T Bell Labs and the MIT Media Lab, wrote Artificial Unintelligence (2018), and appeared in the documentary Coded Bias.

Questions

What is More than a Glitch about?
Meredith Broussard argues that racism, sexism and ableism in technology are not random bugs but patterns built into systems through biased data and design choices. Using cases from facial recognition, lending, policing, grading and medicine, she shows how automated decisions harm people and makes the case for algorithmic auditing and accountability.
Do you need a technical background to read More than a Glitch?
No. Broussard explains machine learning concepts such as training data and black-box models in plain language, using graphs and stories rather than equations. Reviewers have described it as accessible to students, policymakers and newcomers, while practitioners will still find useful case material and arguments for auditing the systems they build.
Is More than a Glitch worth reading in 2026?
Yes, for readers who need to evaluate automated decisions in hiring, lending, health care or public services. Its examples mostly predate the generative AI boom, but its core questions about training data, accountability and auditing apply directly to today's language models and are central to current AI regulation.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.